I. Field of the Invention
This invention relates to the injection molding of plastic materials, and more particularly to the injection molding of products with precisely defined surfaces, such as those which are required to meet optical standards.
While the invention is particularly desirable for products with optical surfaces standards, it can be used generally for the molding of thermosetting and thermoplastic materials.
Optical quality products are in wide use. They are used generally for assays where test substances are subjected to examination by electromagnetic radiation, including visible light. Optical products are also used for instruments, such as microscopes and ophthalmic devices.
Devices that require optical surfaces originally were prepared by grinding glass members. Such products now increasingly employ plastics to expedite manufacture and reduce cost. In general, the demand for plastic optical products is now considerably greater than for glass.
The shift from glass to plastic has occurred primarily because plastic is lighter and often has superior qualities. In addition, protective coatings to provide scratch and abrasion resistance for plastics have become available. Plastic also comes in a wide range of gradient-density tints and colors.
Of the many advantages exhibited by plastics, their relatively light weight and durability have proved to be significant. For optical surfaces, the lens thicknesses are the same for glass and plastic. Consequently, the reduced density of plastic produces a product that is of lighter weight.
The reduction in weight and density is particularly important when high powered surfaces are required, or when large optical surfaces are needed.
Previously, devices with large optical surfaces, particularly those of high power, were typically manufactured by the casting of thermoset resins, for example acrylics that were peroxide cured. However, the availability of polycarbonates and related thermoplastics permits the replacement of cast thermoset plastics. This is because modern polycarbonates have low densities and high refractive indices. For the same optical thicknesses, polycarbonates have an even lower weight than cast plastics, and far lower than glass.
Additionally, since polycarbonates have great impact strength and breakage resistance, they permit the production of relatively thin optical members. Moreover, coatings for polycarbonates are available to provide abrasion resistance. Polycarbonates are particularly suitable for products with "single" optical surfaces, i.e., those with frontal convex and/or backside concave surfaces.
Optical surfaces are defined by two measures of the ray bending power of light or other waves. Spherical power produces magnification and/or reduction, while cylindrical power produces astigmatic corrections. The units of corrective power are in diopters. It often is desirable to have product available with a spherical power in the range from +4 (magnification) to -6 (reduction) diopters, and a cylindrical power in the range from 0 to +2 diopters. Within this range, a volume-frequency distribution can be plotted, centered at zero power. There is reduced frequency in the plot as spherical or cylindrical power increases or decreases.
To be competetive, injection molded products require high yields, with a reduction in scrap and secondary operations, such as trimming.
Additionally, it is desirable to run optical surfaces of differing powers at the same time, without sacrificing productivity, quality or yields. A four-cavity moldset, for example, quadruples the productivity of a particular molding machine. Two of the cavities can be used to mold common spherical and cylindrical power combinations. The remining cavities can be used for less common surfaces.
An illustrative optical surface is found on an optical disk for the laser reading and storage of information. Optical disks for video respond to analog signaling, while compact digital disks are for audio signals. There also is a wide range of computer program disks for information and data storage. These include the CD/ROM (Compact Disk/Read-Only Memory) which is irreversibly encoded with program information, DRAW (Disk Read And Write, i.e. "user write once") and EDRAW (Erasible, i.e. "by the user", Disk Read And Write).
Many disks are encoded during molding by a "stamper", which forms a face surface in the mold cavity. The digital information is represented on the stamper by a spiral of tiny projections, which, in turn, form indentations in the plastic molded disk. A typical indentation has a depth of 0.1 micron and a length of 1-3.3 microns, with a track pitch of 1.6 microns for a spiral array that extends radially outward.
One requirement of high quality molding is intimate contact of the polymer melt with the stamper, without any voids or premature shrinkage. Contact is maintained from the time the cavity is filled with melt until cooling takes place below the glass-transition temperature of the plastic.
Another requirement for optical products is reduction of internal stresses, i.e. "orientation", within the polymer. Ideally, the molding should be "isotropic", i.e., exhibit the same properties in all directions, so that molded stresses and flow induced orientations are eliminated. Such stresses and orientations produce localized differences in ray bending power. The resultant nonuniformities in refractive index are measured in terms of optical path differences, commonly expressed as "birefringence". Avoidance of birefringence is desired.
With surfaces that employ laser signal reading, any flaw which disrupts or deflects the laser beam causes errors. Other properties which require consideration are percentage of light transmission, percentage haze, and the index of discoloration. Localized flaws include opaque specks or clear areas, such as voids or bubbles, which have different refractive indices and optical bending power than adjacent material. Absolute planarity or flatness are often needed where localized warpage would induce prismatic effects and result in off-axis signal transmissions.
In the molding of many optical surfaces it is necessary to conduct operations in clean or "white" rooms. Such rooms provide particle free environments in the range from Class 1,000 to Class 10. Since workers are the biggest source of contamination, automation of handling and post-molding operations is desirable.
Furthermore, for efficiency, microprocessor or CNC (computer numerical control) control should be used. The molding machines also should have individual moldsets, temperature controllers and hopper dryers. A clean air shower is needed for the clamp open and part removal position, together with robotic part pickers.
While optical molding commonly employs a single cavity mold, that makes inefficient use of clean room floor space, and results in a high captial and equipment fixed cost per part. Consequently, it is desirable to produce optical quality parts with multiple cavity molds.
II. The Prior Art
(a) Straight Injection Molding
Early attempts to make acrylic or polycarbonate optical parts used injection molding with the mold cavity surfaces fixed throughout the molding cycle. This required long cycles, high mold surface temperatures approaching the glass-transition temperature of the plastic, along with high plastication and melt temperatures. Slow, controlled fill rates were followed by high packing pressures, which were held until the completion of gate freeze-off.
Fixed cavity processes employ large gating and runner systems to permit appreciable packing pressure and delivery of material before gate freeze-off occurs. At that time no further transfer of molten polymer occurs. Gate freeze-off in fixed cavity injection presents a problem with surfaces having differing radii of curvature. It is the differences in curvature that produce the necessary ray bending needed for optical surfaces. Differing cross-sectional thicknesses result in non-uniform shrinkage during part formation and subsequent cooling. The thickest sections of optical parts are subject to sink marks or depressions which interrupt an otherwise uniform surface. This results in localized aberrations in ray bending character.
Even when care is taken that the injected polymer conforms to the surface of a fixed mold cavity, once gate freeze-off occurs, that prevents additional packing pressure and material transfer. This usually takes place in the thinnest cross-sectional area of the part, and differential shrinkage begins to occur within the melt. The polymer skin then pulls away from the mold surface, with greatest effect in the thickest cross sections. Pre-release, whether partial or complete, of the molded plastic before the cavity is unlocked and opened, detrimentally affects optical quality. The molded contours no longer provide precision surfaces.
Similar problems occur in the straight injection molding of parts with high aspect ratios, i.e., where there are relatively large surface dimensions and relatively small thicknesses. In those cases, a long length of flow is required through a small cross-sectional orifice of the mold cavity.
(b) Polymer Resins
The most widely used polymers for the molding of parts with optical surfaces are polycarbonates and thermoplastic acrylics, particularly polymethyl methacrylate (PMMA).
Acrylics inherently have better flow at low melt temperatures, as well as low birefringence or polymer disorientation. However, they have relatively high water absorption which results in swelling and warpage, and relatively low creep resistance. Susceptibility to heat distortion make acrylics less desirable, except for products, such as video disks, where parts are cemented together with encased information.
Polycarbonates, on the other hand, can have better performance, but are subject to serious processing limitations. Ordinary grades of polycarbonate have a low melt-flow index range, but higher melt-flow grades are available.
Even with high flow grades, the straight injection of polycarbonate causes high birefringence. This is because the mold cavity has fixed dimensions which do not change during the molding cycle, and exceed the finished part by a shrinkage compensation factor.
Polycarbonates are in amorphous chains that form random coils when in a relaxed state. When polycarbonate melt is forced through a restrictive flow path, or orifice, by high injection pressures the polymer distorts from stretching and shearing, realigning the polymer chains so that they are parallel to one another. This is believed to create severe anisotropy, i.e., nonuniformity. The incoming melt front can be regarded as a dynamically stretching zone of molten polymer. In this frontal zone, disorientation is caused by the shear of one polymer layer over another. This is a result of unavoidable velocity differences because the center of flow is faster than at the edges. In the resulting velocity profile, the lowest velocities are at the mold surface, and the highest velocity is at the center. A slowly moving melt front, at low pressure, produces a front that is less distorted and less stressed.
In straight injection molding of polycarbonates, injection is at the highest speed of the hottest, most fluid, polycarbonate melt into the narrow constrictions of a fixed cavity mold.
Elaborate plastication is needed to provide the hottest melt without catastrophic degradation in straight injection molding. Being less viscous, a hot melt provides less internal shear and slower freeze. This allows more time for melt relaxation after flow ceases, and before solidification.
Such plastication can use starved feeding or a reduced sized barrel/screw combination. This minimizes the residence time of the polycarbonate polymer in the injection plastication unit, since high melt temperatures are required. Some plastications cause high shearing of the melt and suffer more polymer degradation.
The balance between degradation flaws--from a hot plastication melt--and high disorientation--from a fast fill rate into a high aspect ratio and restrictive mold cavity, creates a narrow "process window". This has made straight injection suitable only for single cavity molding. Multiple cavity straight injection would result in cavity imbalance.
Another difficulty with straight injection is that the contents of the mold cavity gradually shrink during cooling. This causes the part to pull away from the mold surfaces. Premature release can produce differential warpage or imprecise replication of surface contour patterns. Straight injection uses high injection pressures to maintain cavity pressure until gate freeze-off occurs. However, this application of pressure also causes reextrusion or cold-flow of the increasingly viscous polymer core within the fixed dimensions of the mold cavity. Such forcible redistribution of the partially-solidifying melt creates internal stresses resulting in birefringence.
(c) Injection/Compression Molding
To overcome the difficulties associated with straight injection, resort has been made to mold cavity compression after injection. There are three types: (1) clamping injection/compression, where compression is by platen motion; (2) auxilary component injection/compression, where there is full machine clamping with no platen motion, and mold-cavity compression is by auxilary components internal to the moldset; and (3) clamping, and auxiliary component injection/compression, where mold cavity compression is by clamping and auxiliary component motion.
(C-1) Martin U.S. Pat. No. 2,938,232
As disclosed in Martin U.S. Pat. No. 2,938,232 ("Martin '252") for toggle-clamp injection molding, issued May 31, 1960 and known as a "sandwich press", the mold platens and mold halves are brought together until a predetermined air gap is present at the parting line. At that point, a low pressure, low velocity injection fill begins.
After injection is completed and the molten polymer mass has cooled for a predetermined interval, the machine commences closure of the movable platen. This mechanically seals the mold cavity and its partially solidified contents with zero-clearance at the parting line. The mold halves are locked for the duration of the molding cycle at a predetermined clamp pressure. The partially solidified polymer mass is compressed by the amount of the air gap that existed at the parting line when injection started. By eliminating the air gap, the volume of the cavity and runner system is proportionately affected, resulting in compressive forces exerted upon the partially solidified polymer and causing a reorientation and re-flow. Under clamp induced compressive force, the mold cavity contents continue cooling and solidifying, eventually reaching a temperature sufficiently below glass-transition temperature that the molded part may be ejected without optical distortion.
The result is clamp induced "coining" which offers advantages over straight injection. Successful coining is a function of initial injection pressure and fill rate, air gap dimensions, the timing interval between injection and compression, and the magnitude of the final clamping forces.
Control over injection pressure and fill rate, along with timing are critical. In order to prevent molten polymer from spilling outside the mold cavity, the injected melt must form a surface skin and partially solidify. Otherwise, molten polymer spills into the air-gap and necessitates trimming of the molded part.
If the melt has solidified excessively, compression at ultimate clamping pressures can cause deformation at the parting line and damage the moldset. The cooling interval is critical to achieving acceptable yields. If the melt is not sufficiently solidified at its most constrictive point, partially molten polymer can be extruded out of the cavity and into the runner system. This can result in an underfilled and underpacked part with badly distorted surfaces. However, if compression is delayed too long, too much polymer solidification will occur when the compressive force is initiated. This results in forceable reorientation of the polymer and "cold working" of the plastic, producing birefringence and undesirable molded-in stresses.
Bartholdsten U.S. Pat. No. 4,409,169
To alleviate these problems of Martin '232, Bartholdsten et al U.S. Pat. No. 4,409,169 teaches a slow, low-pressure injection of an oversized shot into a mold that is partially-open at the parting line, followed by deliberate melt cooling, viscosity thickening and a short pressing stroke to squeeze from the reduced mold cavity volume the partially cooled and viscous excess plastic. As pressing continues to the fully closed parting line position, radially extruded overflow is pinched. Full clamping is maintained for shrinkage compensation and avoidance of prerelease.
Matsuda U.S. Pat. Nos. 4,442,061 and 4,519,763
Another clamp induced coining process is disclosed by Matsuda et al in U.S. Pat. Nos. 4,442,061 and 4,519,763. Melt is injected into a slightly opened moldset and cooled until fully solidified. The melt is then reheated uniformly above the melt temperature, at which point a clamp actuated compressive stroke is delivered and maintained throughout a second cooling cycle.
(C-2) Auxiliary Injection/Compression
Another type of injection/compression molding makes use of auxiliary components, such as springs or cylinders to apply compressive force to internal and opposing mold surfaces. The primary difference over clamping injection/compression is that mold compression is provided by a stroke producing element, "whereas mold compression in auxiliary component" molding is provided by auxiliary springs or hydraulic cylinders. Furthermore, clamping injection/compression is sequenced and coordinated by process control, while auxiliary component compression is controlled by self-action, like springs, or separately by timers.
A further distinction is that auxiliary component compression does not employ the motion of a movable platen to provide compressive forces to reduce cavity volume. Instead the mold is fully clamped with no relative motion of the clamp plates, or of fixed and movable platens, during the injection fill, cavity reduction compression, or cooling.
Examples of auxiliary component injection/compression molding are discussed below.
Johnson U.S. Pat. No. 2,443,286
In Johnson U.S. Pat. No. 2,443,286, issued Jun. 22, 1948, spring loaded, movable dies are employed within the moldset. This creates a variable volume mold cavity, but relies upon high internal polymer melt pressure to spread the movable dies against resisting spring pressure. In order to apply a sufficiently great compressive force to the solidifying contents, substantial spring forces are needed. However, the greater the spring force, the greater the injection pressure needed to compress the springs during variable cavity fill. The greater the injection pressure, the greater the degree of molded-in stress and unsatisfactory birefringence. This type of process generally is limited to production of weak optics with small surfaces and limited thickness.
Weber U.S. Pat. Nos. 4,008,031 and 4,091,057
Another auxiliary component process is disclosed in Weber U.S. Pat. Nos. 4,008,031 and 4,091,057. A variable volume cavity is formed by injection melt and pressure induced rearward deflection of at least one movable die. After an interval, forward displacement results in compression under the driving force of an auxiliary hydraulic cylinder mounted in a one-to-one relationship with the movable die. Flow ports are provided for excess, increasingly viscous and partially cooled injected polymer melt which is extruded from the cavity under compressive forces.
Weber teaches slow mold fill, and, as with conventional clamp induced coining, relies upon a preset lapse of time between completion of injection fill and commencement of compressive pressure. Accordingly, Weber is faced with the problems of premature compression, i.e., inadequate solidification, or delayed compression, i.e., late solidification.
In addition, Weber can produce inconsistent parts with variable thicknesses. Depending upon the timing interval, the travel of the movable die is controlled by the length of time elapsing after molten plastic enters the variable cavity and pressure is applied to the movable die. The final volume of the cavity also is controlled by the time elapsing after molten plastic enters the variable cavity, and by the length of time that pressure is applied to the movable die. The result is product variation within the same production run, and thickness variations.
Moreover, when Weber employs a two cavity mold the compression of each cavity is controlled by a separate and independent hydraulic cylinder. Consequently, the two cavities are not simultaneously acted upon by a common component. The larger the number of cavities, the larger the expected variations.
Laliberte U.S. Pat. No. 4,364,878
Another auxiliary component process is disclosed by Laliberte in U.S. Pat. No. 4,364,878. Laliberte includes a movable die coupled to an auxiliary hydraulic cylinder. After the mold is closed under clamp pressure, the mating die parts are spread apart. A precise, volumetrically metered shot that is just adequate to fill the fully-compressed mold cavity is then injected. This control of shot size allows compression without displacement of partly solidified melt out of the mold cavity through an overflow port. The result is greater control over part thickness, eliminating scrap waste and trimming.
However, Laliberte is limited to one-cavity production by reliance upon precisely metered melt, corresponding one-to-one with the injected melt. In addition, there is dependence upon an individually controlled and sequenceable hydraulic cylinder in a one-to-one motion relationship with a variable volume cavity.
While auxiliary component processes have to some degree been useful in molding optical surfaces, they cannot be applied generally.
Compressive forces for auxiliary component molding are much less than those available through clamp actuated coining. This limitation is particularly troublesome for optical surfaces with large projected areas and the necessity for initimate contact with the melt.
(C-3) Clamp and Auxiliary Component Injection/Compression Maus & Galic U.S. Pat. No. 4,828,769
Another prior art technique with clamp and auxiliary component injection/compression is disclosed in U.S. Pat. No. 4,828,769 which issued May 8, 1989 to Steven M. Maus and George J. Galic for injection/compression molding. According to this teaching, an article is formed from a plasticized thermoplastic resin using an injection molding machine in which an opposing pair of mold inserts are initially separated to form a pre-enlarged cavity.
A mass of plasticized resin, slightly larger than the volume of the article to be formed, is injected into the mold cavity. The main clamp force of the injection molding machine is initiated, shortly before completion of the injection to overcome inertial effects. After the completion of injection, the clamping reduces the volume of the closed mold cavity in order to redistribute the resin. The main clamp force is applied until a final clamp lock position is reached.
In addition, the molding machine has first and second mold platens, first and second parting line mold plates, a plurality of parting line. The first and second mold plates, and the first and second mold inserts, are respectivly commonly supported by the first and second mold platens. The mold plates initially are urged together to eliminate any parting line air gap.